Failure must always be an option
I've been scanning Twitter posts and other blog sites and I see that many commentators believe that acceptance of failure and "failing forward faster" is imperative to innovation, and for the most part I agree. What's missing is a deeper investigation into the issues surrounding "failure". How does an organization "accept" some failure when FINAO in most cases? You know a subject is getting some attention when the cartoonists have something to say about it.
The problem with failure - either accepting it or rejecting it - is in defining it. Today, in most firms, failure can be defined in using these criteria:
- No one wants to fail because it reflects badly on them and their careers
- Failure is a one time deal - it's cataclysmic and ends badly
- Failure is an end point - a terminal destination
- Failure is ridiculed, mocked and reviled. We tell cautionary stories about previous failures
- The culture rewards success and ridicules failure.
- Failure is costly
- Failure is unpredictable
When you hear people talking about the importance of failure, or when you read people writing about the importance of failure, they are correct but incomplete. We all understand that failure is important to innovation, but we need to understand how to create the conditions in which failure is acceptable. Clearly, if failure is considered a worthless, terminal condition then it can never be acceptable. However, if we can establish some of the characteristics I've listed below for failure in an organization, then we may be able to create meaningful failure:
- Failure is contained and inexpensive
- Failure is considered a likely outcome
- Failure is not an end point and isn't terminal to the effort
- Failure is considered part of the process of learning and developing
- Failure is predictable - in fact we assert the likelihood of failure
- Because it involves learning, advancement and was predicted, failure is accepted or celebrated
Strangely, the situation I've described is what we'd expect in an R&D lab, but don't plan for or tolerate anywhere else in most organizations. Controlled experiments that advance our knowledge and prepare the first to make better decisions leading to better products - that's valuable. Perhaps we just need to reframe our language. Rather than "failing forward faster" we need "rapid, controlled experimentation".
No matter how many times innovation experts tell you that failure is important and must be part of the process, that doesn't really matter until someone tells you how to make failure more acceptable in your organization, so that you can be allowed to fail occasionally. What managers hate more than failure is uncertainty, unexpected costs and risks. If you can demonstrate that your experiments remove uncertainty, reduce costs and risks, controlled failure will be tolerated or even accepted. Heck, you may even win an award for it.